.Buildings.Fluid.Chillers.ElectricReformulatedEIR

Information

Model of an electric chiller, based on the model by Hydeman et al. (2002) that has been developed in the CoolTools project and that is implemented in EnergyPlus as the model Chiller:Electric:ReformulatedEIR. This empirical model is similar to Buildings.Fluid.Chillers.ElectricEIR. The difference is that to compute the performance, this model uses the condenser leaving temperature instead of the entering temperature, and it uses a bicubic polynomial to compute the part load performance.

This model uses three functions to predict capacity and power consumption:

• A biquadratic function is used to predict cooling capacity as a function of condenser leaving and evaporator leaving fluid temperature.
• A bicubic function is used to predict power input to cooling capacity ratio as a function of condenser leaving temperature and part load ratio.
• A biquadratic functions is used to predict power input to cooling capacity ratio as a function of condenser leaving and evaporator leaving fluid temperature.

These curves are stored in the data record per and are available from Buildings.Fluid.Chillers.Data.ElectricReformulatedEIR. Additional performance curves can be developed using two available techniques (Hydeman and Gillespie, 2002). The first technique is called the Least-squares Linear Regression method and is used when sufficient performance data exist to employ standard least-square linear regression techniques. The second technique is called Reference Curve Method and is used when insufficient performance data exist to apply linear regression techniques. A detailed description of both techniques can be found in Hydeman and Gillespie (2002).

The model takes as an input the set point for the leaving chilled water temperature, which is met if the chiller has sufficient capacity. Thus, the model has a built-in, ideal temperature control. The model has three tests on the part load ratio and the cycling ratio:

1. The test

     PLR1 =min(QEva_flow_set/QEva_flow_ava, per.PLRMax);
   

   ensures that the chiller capacity does not exceed the chiller capacity specified by the parameter per.PLRMax.
2. The test

     CR = min(PLR1/per.PRLMin, 1.0);
   

   computes a cycling ratio. This ratio expresses the fraction of time that a chiller would run if it were to cycle because its load is smaller than the minimal load at which it can operate. Note that this model continuously operates even if the part load ratio is below the minimum part load ratio. Its leaving evaporator and condenser temperature can therefore be considered as an average temperature between the modes where the compressor is off and on.
3. The test

     PLR2 = max(per.PLRMinUnl, PLR1);
   

   computes the part load ratio of the compressor. The assumption is that for a part load ratio below per.PLRMinUnl, the chiller uses hot gas bypass to reduce the capacity, while the compressor power draw does not change.

The electric power only contains the power for the compressor, but not any power for pumps or fans.

The model can be parametrized to compute a transient or steady-state response. The transient response of the boiler is computed using a first order differential equation for the evaporator and condenser fluid volumes. The chiller outlet temperatures are equal to the temperatures of these lumped volumes.

References

• Hydeman, M., N. Webb, P. Sreedharan, and S. Blanc. 2002. Development and Testing of a Reformulated Regression-Based Electric Chiller Model. ASHRAE Transactions, HI-02-18-2.
• Hydeman, M. and K.L. Gillespie. 2002. Tools and Techniques to Calibrate Electric Chiller Component Models. ASHRAE Transactions, AC-02-9-1.

Revisions

• March 12, 2015, by Michael Wetter:
  Refactored model to make it once continuously differentiable. This is for issue 373.
• Jan. 9, 2011, by Michael Wetter:
  Added input signal to switch chiller off.
• September 17, 2010, by Michael Wetter:
  First implementation.